Laparoscopic versus open celiac ganglionectomy in patients with median arcuate ligament syndrome

Transcription

1 From the Society for Clinical Vascular Surgery Laparoscopic versus open celiac ganglionectomy in patients with median arcuate ligament syndrome Allan W. Tulloch, MD, Juan Carlos Jimenez, MD, Peter F. Lawrence, MD, Erik P. Dutson, MD, Wesley S. Moore, MD, David A. Rigberg, MD, Brian G. Derubertis, MD, and William J. Quinones-Baldrich, MD, Los Angeles, Calif Objective: Median arcuate ligament syndrome (MALS) is a rare disorder characterized by abdominal pain and compression of the celiac artery. Traditional management consists of open MAL division, with or without arterial reconstruction. We present our outcomes using a laparoscopic approach and compare them to patients treated with open MAL division during the same period. Methods: A retrospective medical records review of all patients with MALS treated at the University of California Los Angeles from January 1999 to 2009 was performed. Results: Fourteen patients with MALS were treated. All patients underwent an extensive preoperative gastrointestinal (GI) workup with 10 undergoing attempted laparoscopic division of the MAL and celiac ganglion (laparoscopic ganglionectomy [LG]). Two intraoperative conversions were performed for bleeding. Six patients were treated in the open surgery group (open ganglionectomy [OG]). There were no deaths or reoperations in either group. Median time to feeding was 1.0 vs 2.8 days (P <.05) in the LG and OG groups, respectively, which was statistically significant. Median length of hospitalization was also significantly lower in the LG group compared with the OG group (2.3 vs 7.0 days; P <.05). Eight patients had LG (100%) and 5 patients had OG (83%) and had immediate symptom resolution (postoperative day 1). Three patients with recurrent symptoms after LG underwent angiography demonstrating persistent celiac stenosis, then an angioplasty was performed. Median follow-up was 14.0 months (2-65 months) for all patients. Three patients who received LG (38%) and 3 patients who received OG (50%) had persistent pain at last follow-up. Six patients in the OG group (100%) and 7 patients in the LG group (88%) had ceased taking chronic oral narcotics at their last follow-up visit. Conclusion: Both laparoscopic and open MAL division and celiac ganglionectomy can be safely performed with minimal patient morbidity and mortality. Late recurrence is frequently seen; however, this seems to be milder than the presenting symptoms. The laparoscopic approach results in avoidance of laparotomy and was associated with shorter inpatient hospitalization and decreased time to feeding in our study. Optimal patient selection and prediction of clinical response in these patients remains a challenge. (J Vasc Surg 2010;52: ) Median arcuate ligament syndrome (MALS) was first described in Also referred to as celiac artery (CA) compression syndrome and Dunbar s syndrome, it is associated with chronic, severe abdominal pain and extrinsic compression of the CA by the median arcuate ligament (MAL). Other symptoms can include nausea, vomiting, diarrhea, and weight loss. Because the pathophysiology of this disorder remains unknown, its existence as a clinical disorder has been debated leading to significant controversy regarding whether it should be treated and its optimal management. Adding to the controversy is the variable presentation of patients with MALS, with symptoms ranging from mild to incapacitating. From the Division of Vascular Surgery, University of California Los Angeles School of Medicine. Competition of interest: none. Presented at the Thirty-eighth Annual Symposium of the Society for Clinical Vascular Surgery, April 7-10, 2010, Scottsdale, Ariz. Reprint requests: Juan Carlos Jimenez, MD, 200 Medical Plaza Ste 510-6, Los Angeles, CA ( jcjimenez@mednet.ucla.edu). The editors and reviewers of this article have no relevant financial relationships to disclose per the JVS policy that requires reviewers to decline review of any manuscript for which they may have a competition of interest /$36.00 Copyright 2010 by the Society for Vascular Surgery. doi: /j.jvs Traditional management of MALS has consisted of laparotomy through an upper abdominal incision, open division of the MAL, and resection of associated periarterial neural tissue comprising the celiac plexus or ganglion. Restoration of CA patency may be performed either through direct arterial reconstruction or CA bypass. More recently, combined techniques of laparoscopic lysis of the MAL with endovascular angioplasty and/or stenting have been described with minimal patient morbidity. Development of advanced endovascular techniques have led to balloon angioplasty and CA stenting replacing more invasive revascularization procedures for treatment of residual luminal stenoses. To characterize these changes in a therapeutic approach, we prospectively evaluated all patients who underwent both laparoscopic and open surgical treatment for MALS over a 10-year period at a tertiary care hospital and report our patient outcomes. METHODS A retrospective medical records review of all patients undergoing laparoscopic and surgical treatment for MALS from 1999 through 2009 was performed. Approval was obtained by the institutional review board at the University of California Los Angeles for protection of research subjects. A patient database was maintained using Mi- 1283

2 1284 Tulloch et al JOURNAL OF VASCULAR SURGERY November 2010 Table I. Patients treated with laparoscopic median arcuate ligament release Patient Age Gender Year treated Presenting symptoms Comorbidities A 42 F 2004 Persistent pain, worse postprandial, nausea, vomiting, constipation Lupus, supraventricular tachycardia, rheumatoid arthritis, giant-cell arteritis, migraines GERD B 20 F 2006 Postprandial pain, nausea, vomiting C 48 M 2007 Persistent pain, worse postprandial Morbid obesity, hepatic steatosis, peptic ulcer disease D 34 F 2007 Postprandial pain, nausea, vomiting, diarrhea Supraventricular tachycardia, GERD, hepatitis, bulimia, seizure disorder E 19 F 2007 Persistent pain, worse postprandial, positional pain, nausea Pancreatitis F 74 F 2008 Pain persistent, worse postprandial Hypertension, myocardial infarction, breast cancer with mastectomy and radiation G 81 F 2008 Persistent pain, worse postprandial, nausea and vomiting Hypertension, GERD H 46 F 2009 Postprandial pain, nausea, vomiting, diarrhea Hypertensive cardiomyopathy, CVA, oxygen dependence CVA, Cardiovascular accident; EGD, esophagogastroduodenoscopy; ERCP, endoscopically retrograde cholangiopancreatography; F, female; GERD, gastroesophageal reflux disease; GI, gastrointestinal; HIDA, hepatobiliary iminodiacetic acid; M, male; N, no; NA, not applicable; SBFT, small bowel follow through; UGI, upper gastrointestinal; Y, yes. crosoft Excel (Microsoft, Redmond, Wash). Gender, age, presenting symptoms, and comorbidities were reported. Prior gastrointestinal (GI) workup was documented and the outcome of the vascular surgical consultation. Patients with symptoms of chronic abdominal pain, nausea, vomiting, diarrhea, and weight loss underwent preoperative imaging with duplex ultrasonography scan, computed tomography (CT), magnetic resonance angiography, or visceral angiography to confirm the presence of a CA stenosis. Once the diagnosis of MALS was made, patients underwent either laparoscopic ganglionectomy (LG) or open lysis of the MAL (open ganglionectomy [OG]; with complete celiac ganglionectomy performed at the discretion of the attending surgeon). Intraoperative duplex ultrasound scan and angiography were not routinely used in either operative group, and endovascular revascularization was reserved for patients with persistent postoperative symptoms after MAL division. Estimated blood loss and operative times were obtained from the intraoperative anesthesia and operating room records. Statistical analysis was performed where applicable using Excel (Microsoft, Redmond, Wash) and SPSS (SPSS Inc, Chicago, Ill). For normally distributed values, a two-tailed t test was used. The Mann-Whitney U test was used when appropriate for nonparametric variables. Statistical significance was assigned to values at the P.05 level. Laparoscopic technique. The patients were placed supine on the operating table and administered general anesthesia. An optical 5-mm Trocar was used to enter the abdomen midway between the xiphoid and the umbilicus. Additional 5-mm working Trocars were placed under videoscopic guidance after first insufflating the abdomen to 15 mm Hg with carbon dioxide. The typical number of 5 mm laparoscopic ports used was four, including a subxiphoid port, an umbilical port, and two lateral ports. In the early experience, a liver retractor was placed and the right crus of the diaphragm was identified along with the caudate lobe of the liver and the inferior vena cava. The peritoneal line anterior to the right crus was then opened using a harmonic scalpel. Care was taken to identify the esophagus and both anterior and posterior vagal trunks. The stomach was placed on left lateral traction, exposing the right crus. The pars flaccida was divided to open up the area, and dissection was undertaken in this area inferiorly in a caudate direction until the crossing insertion of the left crus was encountered inferiorly. The splenic artery was identified along with the

3 JOURNAL OF VASCULAR SURGERY Volume 52, Number 5 Tulloch et al 1285 Table I. Continued. Prior invasive GI exams and operations Preoperative weight loss (pounds) Follow-up Recurrence of pain Time to pain recurrence Persistent pain at last follow-up Need for continued narcotics Postoperative associated procedures Sigmoidoscopy, colonoscopy, total abdominal hysterectomy Y 6 Y Y Celiac artery angioplasty 4 celiac artery stent 1 EGD, colonoscopy, HIDA scan Y 4 N N Celiac artery angioplasty/ stent 1 EGD, colonoscopy, capsule 0 14 N NA N N None endoscopy, diagnostic laparoscopy 2, liver biopsy Enteroclysis, colonoscopy, ERCP with sphincterotomy, laparoscopic cholecystectomy, appendectomy, laparoscopic pelvic adhesolysis, liver biopsy 5 26 Y 3 N N Visceral angiography 1 without intervention EGD, ERCP 3, sphincterotomy, pancreatic duct stenting, HIDA, colonoscopy, visceral angiography, endoscopic ultrasound, exploratory pelvic laparoscopy, celiac ganglion block EGD, 24 hr ph probe, exploratory laparotomy with adhesolysis, visceral angiography, hysterectomy, cholecystectomy EGD, colonoscopy, ERCP, endoscopic ultrasound, visceral angiography, prior celiac angioplasty/stent 1 EGD, UGI with SBFT, esophageal manometry Y 1 Y N Celiac artery angioplasty N NA N N None N NA N N None 0 8 N NA N N None left gastric trunk. Hypertrophied and fibrotic crura were frequently encountered. The dissection would be carried down until it was posterior to the superior edge of the pancreas. This technique was modified after the first two or three procedures to avoid transection of the muscular body of either the left or right crus, as problems with resultant gastroesophageal reflux were both anticipated and seen. To avoid cutting these muscles as a part of the dissection to gain exposure to the MAL, a more inferior approach has been adopted. This consists of dividing the gastrocolic ligament widely inferior to the gastroepiploic artery, and placing the liver retractor under the posterior gastric wall, elevating the stomach anterior and superiorly to expose the superior border of the pancreas. The dissection then follows the branches of the celiac trunk back to the CA until the MAL is encountered. By taking this approach, the crura can be spared division, and there is some thought that the dissection follows a more natural angulation, given the laparoscopic access to the peritoneal space. The overall amount of dissection is decreased via this approach; however, there is some need for an increased level of awareness to protect the celiac branches and the pancreas during the procedure. The entire region was dissected cleanly and the MAL and the celiac axis nerve fibers were carefully identified and divided with the harmonic scalpel down to the origin of the celiac axis. Careful skeletonization of all ligamentous and neural fibers surrounding the CA and the proximal aorta was performed. The artery was visualized after MAL division to assess for any remaining extrinsic defects; however, intraoperative duplex and/or angiography were not routinely used. Conditions for hospital discharge included the ability to tolerate adequate oral intake and good pain control with oral analgesics. Open technique. With the patient under general anesthesia, a curvilinear transverse subcostal or midline incision was performed at the surgeon s discretion. The lesser sac and lesser omentum were opened and the stomach was retracted downward. Dissection was begun along the hepatic artery and continued in the perivascular plane toward the confluence of the splenic and left gastric arteries to the main trunk of the CA. After dissecting out the branches of the celiac axis, the arcuate ligament and all of its fibers were then divided. All surrounding ligamentous and ganglionic tissue was sharply dissected and removed down to the aorta and circumferentially mobilized. The abdominal wall was closed in layers. Conditions for hospital discharge were the same as those described for the LG. Postoperative treatment protocol. All patients were evaluated postoperatively as inpatients and in the outpa-

4 1286 Tulloch et al JOURNAL OF VASCULAR SURGERY November 2010 Table II. Patients treated with open lysis of the median arcuate ligament Patient Age Gender Year treated Presenting symptoms Comorbidities I 22 F 2004 Postprandial pain None J 16 M 2004 Postprandial pain, nausea, vomiting Pectus carinatum K 53 F 2004 Persistent pain, worse postprandial, positional pain, nausea, vomiting Atrial fibrillation, asthma, breast cancer with bilateral mastectomies and radiation L 18 F 2007 Persistent pain, nausea, positional Heart transplant for idiopathic dilated cardiomyopathy, pancreatitis pain M 74 F 2009 Postprandial pain, nausea, vomiting, diarrhea Uterine cancer, breast cancer, lung cancer s/p TAH, BLSOO, lung resection (LLL), R MRM, osteopenia, DJD, small bowel intussusception from lipoma s/p jejunal resection, smoking, hypothyroidism N 78 F 2009 Persistent pain, positional pain Atrial fibrillation, osteoporosis, vulvar cancer s/p TAH, s/p rectocele repair BLSO, Bilateral salpingoophorectomy; DJD, degenerative joint disease; EGD, esophagogastroduodenoscopy; F, female; F/U, follow-up; LLL, left lower lobe; M, male; N, no; NA, not applicable; R MRM, modified radical mastectomy; SBFT, small bowel follow through; s/p, status post; TAH, total abdominal hysterectomy; UGI, upper gastrointestinal series; Y, yes. tient clinic for persistent abdominal pain. Asymptomatic patients were monitored clinically and imaging studies were reserved for patients with persistent abdominal symptoms. Patients with complaints of abdominal pain, nausea, vomiting, or weight loss underwent visceral angiography to evaluate the CA for residual luminal stenosis. Primary CA stenting was not performed. Patients with residual luminal stenoses of greater than 30% and/or pressure differential of greater than 10 mm Hg underwent balloon angioplasty of the CA. Celiac stenting was reserved for patients with residual stenoses after balloon angioplasty with the criteria described above. All patients who underwent endovascular intervention of the CA were monitored postoperatively using duplex ultrasound scan, CT, magnetic resonance angiography, or conventional angiography at 6-month intervals. RESULTS Fourteen patients were treated for MALS over a 10- year period ( ). Ten patients underwent attempted laparoscopic MAL lysis and celiac ganglionectomy (LG). Two patients required open conversion for bleeding and were included in the open surgical group (OG). Six patients underwent laparotomy with open lysis of the MAL and celiac ganglionectomy. Preoperative patient characteristics and postoperative results are summarized for each individual patient in Tables I and II. Preoperative results. Patients included 12 women (86%) and 2 men (14%). The mean age in both groups was years. Patients underwent an extensive preoperative GI workup with a mean number of invasive diagnostic and surgical procedures for both groups (Tables I and II). All patients had the majority of their preoperative workup and imaging performed at outside hospitals before vascular surgery referral to our institution. Median preoperative weight loss was 12.5 (0-63 pounds). Pain was described as persistent in 8 patients (57%) and worsened after eating in 11 patients (79%). Five patients described the pain as positional (36%) and 9 (64%) presented with associated nausea and vomiting. Two patients in the OG group underwent prior celiac stenting at another institution before MAL lysis and presented with persistent celiac stenosis and abdominal pain. One patient in the OG group underwent balloon angioplasty of the CA (by a different specialist at our institution) before surgical open MAL lysis resulting in acute thrombosis of the main trunk (Fig). We performed an open thrombectomy and patch angioplasty of the celiac and common hepatic arteries, and surgical MAL lysis was performed at this time. All patients in both groups were taking long-term oral narcotic analgesics for chronic abdominal pain. Procedural results. The technical success rate in the LG group was 80%. Twenty percent required conversion for bleeding. Operative conversion was required for acute bleeding from the left gastric artery in 1 patient and the celiac trunk in the other. Both open conversions occurred early in our experience and likely occurred due to subadventitial dissection and excessive thinning of the artery with the harmonic scalpel. Direct suture repair of the left gastric artery was performed during the first conversion, and patch angioplasty was required to repair the celiac trunk arteriotomy during the second. Median operative blood loss in the LG and OG was 50 ml ( ml) and 80 ml ( ml), respectively. Median estimated blood loss in the OG, excluding the 2 patients who were acutely converted, was 10 ml ( ml). These values were not statistically significant (P.05). Median operative times were 220 minutes ( minutes) in the LG group and minutes ( minutes) in the OG group. These values also failed to meet statistical significance (P.05). No patients required antegrade or retrograde CA bypass. There were no postoperative deaths or reoperations in either group. All patients who received LG (8 of 8; 100%) demonstrated immediate postoperative relief of symptoms

5 JOURNAL OF VASCULAR SURGERY Volume 52, Number 5 Tulloch et al 1287 Table II. Continued. Prior invasive GI exams and operations Preoperative weight loss (pounds) F/U Recurrence of pain Time to pain recurrence Persistent pain at last F/U Need for continued narcotics Postoperative associated procedures Visceral angiography 15 5 N NA N N None Visceral angiography 30 1 N NA N N None Visceral angiography Y 42 Y N Celiac artery angioplasty/stent 1 Visceral angiography and celiac 0 37 N NA N N None artery angioplasty Exploratory laparoscopy, EGD, UGI with SBFT Y 1 Y N None EGD, intestinal biopsy, colonoscopy, visceral angiography 0 3 Y 2 Y N None Fig. Acute thrombosis of the celiac trunk and its branches after aggressive balloon angioplasty before median arcuate ligament release. Endovascular intervention of the celiac artery should only be performed after ligament release in patients with median arcuate ligament syndrome. compared with 5 of 6 patients in the OG group (83%). Median time to feeding was significantly faster in the LG group compared to the OG group. These values were 1.0 days in the LG group (patients tolerated feeds on postoperative day 1) vs 2.8 days (2-4 days) in the OG group (P.05). Length of hospitalization was also significantly shorter in the LG group compared with the OG group and was 2.0 days (2-3 days) vs 7.0 days (5-8 days), respectively (P.05). Postoperative results. Median follow-up for all patients treated was 14.0 months (2-65 months). Median follow-up in the LG and OG groups was 14.0 months (2-65 months) and 10.0 months (2-48 months), respectively. All patients (100%) in the LG group exhibited immediate postoperative pain relief (defined as postoperative day 1). Four patients (50%) in the LG group developed early recurrence of symptoms (defined as less than 6 months) and underwent visceral angiography after MAL lysis at a time of 3.5 months. Three of these 4 patients were found to have hemodynamically significant CA stenoses and underwent balloon angioplasty. Two required stent placement for persistent CA stenosis after balloon angioplasty. One patient in the LG group with recurrent abdominal pain underwent postoperative angiography, which revealed no evidence of celiac stenosis, and no intervention was performed. Two patients treated with celiac stents in the LG group presented with persistent abdominal pain at last follow-up, but reported mild to moderate improvement. One patient continues to experience chronic, unremitting abdominal pain similar to the preoperative presentation. All patients in the LG group with persistent abdominal pain at last follow-up underwent CT angiography demonstrating widely patent CA stents. Four patients in the LG group did not undergo postoperative angiography because they remained symptom-free after MAL lysis. An additional patient with early postoperative pain subsequently resolved spontaneously at 26-month follow-up and was pain-free at last follow-up. Five of 6 patients (83%) in the OG group experienced immediate postoperative pain relief. One patient with severe early postoperative abdominal pain was found to have a splenic infarction from intraoperative embolization and her abdominal pain resolved at 1 month. Three patients (50%) developed recurrent abdominal pain in the OG group despite the presence of a patent CA stent on postoperative CT angiography. Abdominal pain in these patients was subjectively described as milder and less lifestyle limiting than at presentation. One patient who underwent celiac stent placement at an outside institution, before MAL lysis, developed late (defined as greater than 6 months) recurrent symptoms at 42 months postoperatively and was found to

6 1288 Tulloch et al JOURNAL OF VASCULAR SURGERY November 2010 have an occluded stent. She refused further intervention. The median time to recurrence of symptoms in the LG group was 4.0 months (1-6 months) and 2 months (1-42 months) in the OG group. Six patients in the OG group (100%) and 7 patients in the LG group (88%) had ceased taking chronic oral narcotics at their last follow-up visit. DISCUSSION Since its first description by Harjola 1 in 1963, authors have continued to debate the existence of MALS as a clinical syndrome. The variability of patient presentation, the unclear understanding of its pathophysiologic mechanism, and its unpredictable response to treatment have all fueled the discussion. An additional confounding factor includes a significant number of asymptomatic patients with radiographic evidence of CA compression. Park et al 2 performed 400 consecutive celiac arteriograms in asymptomatic individuals undergoing hepatic tumor chemoembolization. The percentage of patients with a greater than 50% luminal CA stenosis and a pressure gradient greater than 10 mm Hg was 7.3% in their study. A landmark study reviewing surgical treatment and long-term follow-up in patients with MALS was performed by Reilly et al. 3 In this series, 51 patients underwent operative treatment for symptomatic CA compression over a 17-year period. Mean follow-up was 9 years. Sixteen patients underwent celiac decompression only and 17 underwent decompression and open CA dilatation. Eighteen patients underwent decompression and vessel reconstruction by either primary reanastomosis or interposition grafting. A postprandial pain pattern, age between 40 and 60 years, and preoperative weight loss of 20 pounds or more were associated with sustained symptom relief. Factors negatively associated with clinical relief included: an atypical pain pattern with periods of remission, psychiatric disorder, alcohol abuse, age greater than 60 years, and weight loss less than 20 pounds. Patients treated with both splanchnic nerve decompression and vessel reconstruction experienced better symptom relief (76%) than patients treated with celiac decompression alone (53%). A stenosed or occluded CA on angiography was noted in 75% of patients who presented with persistent symptoms. Excluding case reports, only two other series have reported their experience with laparoscopic treatment of MALS. To our knowledge, this study is the first direct comparison of consecutive patients with MALS treated using both open and laparoscopic techniques. Roseborough 4 performed laparoscopic MAL lysis in 15 patients over a 5-year period. In his series, 14 of 15 patients (93%) subjectively reported significant improvement at a mean follow-up of 44.2 months. One patient remained with severe refractory symptoms. Four patients required open conversion for bleeding. Despite one case of severe pancreatitis, no other complications or deaths were noted. Baccari et al 5 treated 16 patients with a similar laparoscopic technique over a 7-year period. In this series, 2 patients required open conversion for bleeding. Fourteen patients were noted to be asymptomatic at postoperative follow-up. Two patients with residual celiac stenosis were symptomatic and completely resolved their abdominal pain after revascularization (aortoceliac bypass and percutaneous transluminal angioplasty/stent). Our findings demonstrate that both laparoscopic and open MAL lysis with celiac ganglionectomy can be performed safely with good symptom relief in selected patients. Despite the two open conversions early in our experience and significant patient comorbidities, no major complications or perioperative death occurred in either group. However, patients must be counseled preoperatively that the risk of open conversion is possible with the mean rate of conversion in all three published series approaching 20%. 4,5 All patients in the LG group tolerated a diet on postoperative day 1 and advanced to a regular diet without incident. This contributed to the significantly shorter hospital stay that was also noted in the LG group. These advantages in the LG group along with comparable late recurrence rates suggest that a laparoscopic approach may be the preferred technique in patients with MALS. Currently, all patients diagnosed with MALS at our institution are evaluated initially for laparoscopic MAL release. Poor candidates include patients with multiple prior abdominal operations and intraperitoneal adhesions. We believe preoperative CT and magnetic resonance angiography provide excellent noninvasive alternatives to conventional angiography for visualization extrinsic CA compression by the MAL. Significant long-term symptomatic improvement was noted in the majority of patients treated in both the LG group and OG group in our study. Although routine intraoperative celiac revascularization (open celiac dilatation, patch angioplasty, bypass) after MAL division has been advocated by some authors in earlier studies, 3 we believe that balloon angioplasty and/or stenting are effective endovascular techniques which can be used in the postoperative period in patients with persistent postoperative symptoms. These techniques may avoid the morbidity associated with aortic and CA clamping and the welldocumented effects of visceral ischemia and reperfusion intraoperatively. Recurrence was seen in a substantial number of patients treated in both groups and was seen at a median period of 3 months postoperatively. However, with the exception of 1 patient in the OG group who developed late thrombosis of her celiac stent, all patients with recurrent symptoms in both groups had less severe pain than at their initial presentation, and all had radiographic evidence of a widely patent CA postoperatively. This finding is more consistent with the pathophysiologic mechanism of neurogenic etiology favored by some authors, 6,7 and may be another reason to avoid routine CA revascularization in patients undergoing MAL division. This finding also conflicts with the findings in Reilly et al s 3 article in which the majority of patients with recurrent abdominal pain were found to have a stenosed or occluded CA. The increased prevalence of persistent symptoms in our series also differs from the other two laparoscopic series discussed. 4,5

7 JOURNAL OF VASCULAR SURGERY Volume 52, Number 5 Tulloch et al 1289 Based on the findings of this study, an accurate prediction of response to treatment and optimal patient selection in MALS remains a challenge. All patients underwent an exhaustive, comprehensive preoperative GI workup with positive findings limited to CA compression. Referral to vascular surgery for palliation followed multiple invasive GI procedures and operations in the majority of the patients treated in this series (Tables I and II). We did not perform preoperative inspiratory and expiratory lateral studies on all patients, which may increase the accuracy of MALS diagnosis. However, 2 of 3 patients with recurrent symptoms in the LG group underwent preoperative inspiratory and expiratory lateral aortography demonstrating increased narrowing on expiratory views, a common diagnostic sign of MALS. Gastric tonometry has also been described as a sensitive diagnostic test in patients with MALS to detect changes in gastric mucosal ischemia after CA decompression and may be another method to assist with optimal patient selection This technique was not used in our study. Although no patient in our series had a known psychiatric diagnosis or a history of substance abuse, no formal psychiatric evaluation was performed on our patients. This represents another potential limitation of our study. Reilly et al 3 demonstrated a poor response to treatment in patients with a history of alcohol abuse or a formal psychiatric disorder. We have now begun formally evaluating prospective patients with routine preoperative psychiatric evaluation. CONCLUSION Both laparoscopic and open MAL lysis and celiac ganglionectomy can be safely performed with minimal patient morbidity and mortality. Late but milder recurrence of symptoms is frequently seen after both approaches. The laparoscopic approach results in avoidance of laparotomy and may be associated with shorter inpatient hospitalization and decreased time to feeding. Optimal patient selection and prediction of clinical response remains a challenge. AUTHOR CONTRIBUTIONS Conception and design: JJ, DR Analysis and interpretation: JJ, PL, BD, WM, DR, WQ, ED Data collection: AT, JJ, ED Writing the article: AT, JJ, PL, ED Critical revision of the article: JJ, PL Final approval of the article: JJ Statistical analysis: AT, JJ Obtained funding: JJ, PL Overall responsibility: JJ REFERENCES 1. Harjola PT. A rare obstruction of the celiac artery. Report of a case. Ann Chir Gynaecol Fenn 1963;52: Park CM, Chung JW, Kim HB, Shin SJ, Park JH. Celiac axis stenosis: incidence and etiologies in asymptomatic individuals. Korean J Radiol 2001;2: Reilly LM, Ammar AD, Stoney RJ, Ehrenfeld WK. Late results following operative repair for celiac artery compression syndrome. J Vasc Surg 1985;2: Roseborough GS. Laparoscopic management of celiac artery compression syndrome. J Vasc Surg 2009;50: Baccari P, Civilini E, Dordoni L, Melissano G, Nicoletti R, Chiesa R. Celiac artery compression syndrome managed by laparoscopy. J Vasc Surg 2009;50: Marable SA, Kaplan MF, Beman FM, Molnar W. Celiac compression syndrome. Am J Surg 1968;115: Watson WC, Sadikali F. Celiac axis compression: experience with 20 patients and a critical appraisal of the syndrome. Ann Intern Med 1977;86: Balaban DH, Chen J, Lin Z, Tribble CG, McCallum RW. Median arcuate ligament syndrome: a possible cause of idiopathic gastroparesis. Am J Gastroenterol 1977;92: Mensink PB, van Petersen AS, Kolkman JJ, Otte JA, Huisman AB, Geelkerken RH. Gastric exercise tonometry: the key investigation in patients with suspected celiac artery compression syndrome. J Vasc Surg 2006;44: Faries PL, Narula A, Veith FJ, Pomposelli FB Jr, Marsan BU, LoGerfo FW. The use of gastric tonometry in the assessment of celiac artery compression syndrome. Ann Vasc Surg 2000;14:20-3. Submitted Apr 9, 2010; accepted May 12, 2010.